1. MotivationThis poster addresses anomalous acidity values in UK rainfall, recorded while seeking a homeostatic mechanism in a planetary atmosphere friendly to complex carbon life forms for at least 400M years.

2. Fieldwork( Fig. 1)The initial data was collected over two summer months in 2009 at Reading University’s Meteorological Observation Site, measured in real time by pH with RMS averages recorded at 60-second intervals. Figure 1 shows that acidity spikes are followed by c. 4-hour decay curves after rain had stopped, a phenomenon that was provisionally attributed to `volatile acidity’. The decay curve could be replicated indefinitely by recharging a rain sample with pure carbon dioxide (CO2) and exposing it to fresh air.

3. Henry’s Law equilibrium (Figs 2 ; 3)Assuming provisionally that the acid agent was CO2, the rainfall had been supersaturated, the saturated equilibrium being represented by the baseline value following each decay cycle in Fig. 1. The equilibrium solubility of CO2 in water is conventionally measured at 100.3KPa pressure of the solute gas. However for substantially lower pressures (Earth sea level = c 40Pa) the published data is not linear (Fig 3) . Indeed Kurt Buch (1925) reported a consistent positive deviation at pressures down to 100Pa (Fig 3). His values were on the high side relative to spectrometry in the present project, but his observation is here referred to as the `Buch Effect’.

4. The question? How could rainwater be `supersaturated’ on arrival if it had formed in clouds with CO2 solute pressure around 85% lower? Analytical methods generally are not equipped to contain supersaturation, perhaps explaining why such a transport might go unrecognised. Granat speculated `oversaturated’ CO2 content, but concluded that his laboratory procedures would have eliminated it (1972).

6. A potential mechanism: (Fig. 5) A recent communication to Faraday Discussions adds a temperature dimension to the Buch Effect, suggesting that CO2 solubility tracks the molar volume of water at temperatures towards (and logically beyond) its density-maximum (FD 167 (2014), 462-3).

7. The hypotheses: This poster postulates:1. Volatile acidity measured in Reading was CO2;2. It had diffused into cloud drops supercooled during convection;3. It may be of value in modelling convection.

8. New work - replication of real-time pH analyses using gas-phase NDIR spectrometry (Fig. 4) 7. Real-time monitoring: A 200mm glass-packed column

carries a 0.1 Lpm flow of CO2 -free air, stripping CO2 from a dosed feed of rainwater, the gas phase analysed by Li-Cor 820 and SenseAir K30 NDIR spectrometers;

8. Calibration: A headspace volume of 2.3L air (CO2 -free) is re-cycled through 1mol rainwater for absolute content – headspace analysed by Li-Cor 820 NDIR spectrometer.

9. Initial results and synoptic Met data (Fig. 5) Jointly Figs 1 and 5 confirm that, even with the Buch Effect, the saturated CO2 content of rainwater is insignificant, the bulk CO2 transport being the volatile component. Moreover the spikiness in each case suggests a link between CO2 scavenging and convection.

10. Next stepsIf CO2 is being scavenged, is it on a scale that depletes the atmosphere. Conversations are starting with the universities of Manchester and Reading for assembling an archive of flight data and sonde data on variations in the mol fraction of CO2 associated with convection. In parallel, an archive of 260 Oxford rainwater analyses for CO2, ΔpH and Δconductivity is being studied for any trends in chemical scavenging.

Wider implications: `Volatile acidity’; the `Buch Effect’; and `convective scavenging’; are presented here as physical properties of the water molecule and hence operative in Earth atmosphere since the Carboniferous and before. The findings suggest therefore that some of the CO2 capture presently attributed to photosynthesis and dry deposition is in fact wet deposition. This may be of value in modelling convection and atmospheric homeostasis.

High resolution real-time analyses of precipitation at two UK ground locations:

Dissolved CO2 as convection monitor? Brian Durham, University of Oxford

* Research Associate, RLAHA, University of Oxford Grateful thanks for generous financial assistance: Meteorology Department, University of Reading; Royal Meteorological Society; The Rainfall Organisation; and John Durham. Thanks for advice, support and discussion to: Neil Cape; Stephen Cox; Lennart Granat; Giles Harrison: Brian Howard; John Methven; Mirjam Orvomaa; Christian Pfrang; Mark Pollard; Christoff Salzmann; Mike Stroud; Holger Tost; Geraint Vaughan; Ana Vila Verde; Ray Weiss.

Fig. 2

Solubility of CO2 in pure water against

temperatures at partial pressures c.

450 ppm of CO2 . Values are RMS averages of 60 readings at the

manufacturer’s 3% confidence level.

Fig. 1 Field results 2009 This `spiky’ trace of acidity is a 7-day extract from the 2009 Reading series. Note the `decay curve’ over four hours after each spike, as the 4mL sample well equilibrates after rain had stopped. Following the rain of 18 July the equilibrium acidity (base line) is marginally higher than following other events that week (i.e. non-volatile acidity). Previous analyses (without measures to contain volatility) are likely to have reported only the equilibrium CO2 content (base line).

12/07/2009

13/07/2009

14/07/2009

15/07/2009

16/07/2009

17/07/2009

18/07/2009

19/07/20090

20

40

60

80

100

120

140

160

0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0.9

A week of summer rainfall `volatile acidity', Reading, UK, 2009 (mm rain/[hydrogen ion+] per L)

[H3O+] µ mols per litre

Rainfall (calibrated)

Rain

wat

er a

cidity

conv

erte

d fr

om p

H (=

(10

(̂-pH)

) (µm

ols

[H3O

+] p

er L)

Rain

fall

(ARG

100

tipp

ing

buck

et ca

libra

ted

for 1

m2

rain

co

llect

or) (

mm

rain

fall)

Fig. 5 Preliminary results: On 4 April 2015 Oxford saw the passage of a warm sector (pressure chart 5b) with a series of rain events including that illustrated on the 16:30 radar image 5c, rain tracking south-eastwards (see red arrow). Calibration of stripped CO2 is in progress, i.e. water-flow/air-flow/ completeness of stripping, but the similarity to [H3O+] Fig. 1 is evident.

00:00:0002:00:00

03:59:5905:59:59

07:59:5909:59:59

11:59:5913:59:59

15:59:5917:59:59

19:59:5921:59:59

23:59:59-20

0

20

40

60

80

100

120

140

0

0.02

0.04

0.06

0.08

0.1

0.12

0.14

0.16

Oxford, 4 April 2015: Rainfall against stripped CO2 (CO2 uncalibrated)

CO2 Li-Cor

Rainfall (mm)

Strip

ped

CO2

ppm

(0.1

L air/

12.2

2g ra

inw

ater

per

min

ute)

Rain

fall

rate

(mm

rain

fall)

Water structure: Do changes to the bond angle and bond length of water (density maximum 4 deg C) improve its low temperature capacity for dissolving CO2?

Li-Cor 820NDIR

analyser

Peristaltic pumpViton tubing

Sodalime stack

Strippingcolumn

Water out

Koboldflowmeter

Water-level

setting

Rain collect

or

Experiment 6: CO2 stripper

air

air

-40 -30 -20 -10 0 10 20 30 40 500

0.5

1

1.5

2

2.5

Solubility of CO2 at c. 45 Pa: measured (Faraday Discussions 167, 2014) against calculated (Sander

2014)

Mu mols CO2 measured 2013 (Faraday Discussions 167, 2014)

mu mols CO2 from constant (Sander 2014)

Temperature (degrees C)

Diss

olve

d CO

2 un

der 4

5Pa

CO2

head

spac

e (µ

mol

s CO

2 pe

r mol

wat

er)

0.000.01

0.101.00

10.00100.00

1000.0010000.00

100000.00

1000000.00

0.1

1

10

100

1000

10000

100000

Solubility of CO2 in water against partial pressure at c. 20'C: values available 2013

Wiebe & Gaddy (1940) at 18°C

Morgan and Maass (1931) at 18°C

Novak et al (1961) at 20°C

Carroll, Slupsky and Mather (1991) model at 20°C

Buch (1925) ≈ at 19°C

Henry's Law prediction

Pure Water at ≈ 22°C (new values, provisional)

Pressure/partial pressure of CO2 in headspace (kPa CO2, log 10)

Dis

solv

ed C

O2

(µm

ols

CO2

per m

ol w

ater

, log

10)

CO

2 in

mod

ern

Eart

h tr

opos

pher

e

Fig 3

Log-log axes illustrate reported deviations from the Henry constant in the region of CO2 partial pressures in Earth atmosphere. New values are RMS averages of 60 readings at the manufacturer’s 3% confidence level.

Henry’s Law and the `Buch Effect’

Fig. 4 Schematic CO2 strippers

Left: real time analyserair flow 0.1 L/min; rainwater dosed 17 mL/min;

Right: closed headspace for absolute CO2 values

2.3 L air circulated @ 0.8 L/min; rainwater 18 mL (= 1 mol), air flow directed by four Hamilton loop valves,

optional 250mL flask for low end readings.

Brian Durham